皮肤病识别分割数据集labelme格式9082张19类别

数据集中1/3是原图剩余为增强图片，请仔细查看图片预览。

数据集格式：labelme格式(不包含mask文件，仅仅包含jpg图片和对应的json文件)

图片数量(jpg文件个数)：9082

标注数量(json文件个数)：9082

标注类别数：19

标注类别名称:["acne","eczema","normal skin","basal-cell-carcinoma","Tinea-Corporis","Hemangiome","Darier_s-Disease","Psoriasis","Epidermolysis-Bullosa-Pruriginosa","Impetigo","Molluscum-Contagiosum","Hailey-Hailey-Disease","Leishmanios","lichen","Lupus-Erythematosus-Chronicus-Discoides","Melanoma","nevus","Porokeratosis","Tungiasis"]

每个类别标注的框数：

acne count = 11763

eczema count = 1414

normal skin count = 928

basal-cell-carcinoma count = 563

Tinea-Corporis count = 868

Hemangiome count = 885

Darier_s-Disease count = 786

Psoriasis count = 4427

Epidermolysis-Bullosa-Pruriginosa count = 4359

Impetigo count = 2832

Molluscum-Contagiosum count = 6958

Hailey-Hailey-Disease count = 669

Leishmanios count = 311

lichen count = 11386

Lupus-Erythematosus-Chronicus-Discoides count = 1613

Melanoma count = 967

nevus count = 520

Porokeratosis count = 2432

Tungiasis count = 428

使用标注工具：labelme=5.5.0

所在仓库：firc-dataset

图片分辨率：600x600

标注规则：对类别进行画多边形框polygon

重要说明：可以将数据集用labelme打开编辑，json数据集需自己转成mask或者yolo格式或者coco格式作语义分割或者实例分割

特别声明：本数据集不对训练的模型或者权重文件精度作任何保证

类别翻译：

{

 "acne": "痤疮",

 "eczema": "湿疹",

 "normal skin": "正常皮肤",

 "basal-cell-carcinoma": "基底细胞癌",

 "Tinea-Corporis": "体癣",

 "Hemangiome": "血管瘤",

 "Darier_s-Disease": "达里埃病",

 "Psoriasis": "银屑病",

 "Epidermolysis-Bullosa-Pruriginosa": "大疱性表皮松解症",

 "Impetigo": "脓疱疮",

 "Molluscum-Contagiosum": "传染性软疣",

 "Hailey-Hailey-Disease": "海利-海利病",

 "Leishmanios": "利什曼病",

 "lichen": "苔藓",

 "Lupus-Erythematosus-Chronicus-Discoides": "慢性盘状红斑狼疮",

 "Melanoma": "黑色素瘤",

 "nevus": "痣",

 "Porokeratosis": "汗孔角化症",

 "Tungiasis": "潜蚤病"

}

图片预览：

标注例子：